Search

Changes in the consumption price of aquatic products will affect demand and fishermen's income. The accurate prediction of consumer price index provides important information regarding the aquatic product market. Based on the non-linear and non-smooth characteristics of fishery product price series, this paper innovatively proposes a fishery product price forecasting model that is based on Variational Modal Decomposition and Improved bald eagle search algorithm optimized Long Short Term Memory Network (VMD-IBES-LSTM). Empirical analysis was conducted using fish price data from the Department of Marketing and Informatization of the Ministry of Agriculture and Rural Affairs of China. The proposed model in this study was subsequently compared with common forecasting models such as VMD-LSTM and SSA-LSTM. The research results show that the VMD-IBES-LSTM model that was constructed in this paper has good fitting results and high prediction accuracy, which can better explain the seasonality and trends of the change of China's aquatic product consumer price index, provide a scientific and effective method for relevant management departments and units to predict the aquatic product consumer price, and have a certain reference value for reasonably coping with the fluctuation of China's aquatic product market price. [ABSTRACT FROM AUTHOR]

The authors note that: "Fig. 5 in the published paper appeared incorrectly. The correct figure and the figure caption are provided below. The main message and the interpretation of our paper remain unaffected by this correction." [ABSTRACT FROM AUTHOR]

The notion of long-term memory has received considerable attention in empirical finance. This paper makes two main contributions. First one is, the paper provides evidence of long-term memory dynamics in the equity market of China. An analysis of market patterns in the Chinese market (a typical emerging market) instead of US market (a developed market) will be meaningful because little research on the behaviors of emerging markets has been carried out previously. Second one is, we present a comprehensive research on the long-term memory characteristics in the Chinese stock market returns as well as volatilities. While many empirical results have been obtained on the detection of long-term memory in returns series, very few investigations are focused on the market volatility, though the long-term dependence in volatility may lead to some types of volatility persistence as observed in financial markets and affect volatility forecasts and derivative pricing formulas. By means of using modified rescaled range analysis and Autoregressive Fractally Integrated Moving Average model testing, this study examines the long-term dependence in Chinese stock market returns and volatility. The results show that although the returns themselves contain little serial correlation, the variability of returns has significant long-term dependence. It would be beneficial to encompass long-term memory structure to assess the behavior of stock prices and to research on financial market theory. [ABSTRACT FROM AUTHOR]

This paper presents a hybrid model, named as hybrid deep neural network, which combines convolutional neural network, particle swarm optimization, and gated recurrent unit, termed as convolutional neural network-particle swarm optimization-gated recurrent unit model. The major aims of the model are to perform accurate electricity theft detection and to overcome the issues in the existing models. The issues include overfitting and inability of the models to handle imbalanced data. For this purpose, the electricity consumption data of smart meters is taken from state grid corporation of China. An electric utility company gathers the data from the intelligent antenna-based smart meters installed at the consumers' end. The dataset contains real-time data with missing values and outliers. Therefore, it is first preprocessed to get the refined data followed by feature engineering for selection and extraction of the finest features from the dataset using convolutional neural network. The classification of electricity consumers is performed by dividing them into honest and fraudulent classes using the proposed particle swarm optimization-gated recurrent unit model. The proposed model is evaluated by performing simulations in terms of several performance measures that include accuracy, area under the curve, F 1 -score, recall, and precision. The comparison between the proposed hybrid deep neural network and benchmark models is also performed. The benchmark models include gated recurrent unit, long short term memory, logistic regression, support vector machine, and genetic algorithm-based gated recurrent unit. The results indicate that the proposed hybrid deep neural network model is more efficient in handling class imbalanced issues and performing electricity theft detection. The robustness, accuracy, and generalization of the model are also analyzed in the proposed work. [ABSTRACT FROM AUTHOR]

This paper investigates the expected results of the current COVID-19 outbreak to arrivals of Chinese tourists to the USA and Australia. The growing market share of Chinese tourism and the fact that the county was the first to experience the pandemic make China a suitable proxy for predictions on global tourism. We employ data from the 2003 SARS outbreak to train a deep learning artificial neural network named Long Short Term Memory (LSTM). The neural network is calibrated for the particulars of the current pandemic. Our findings, which are cross-validated using backtesting, suggest that recovery of arrivals to pre-crisis levels can take from 6 to 12 months and this can have significant adverse effects not only on the tourism industry but also on other sectors that interact with it. [ABSTRACT FROM AUTHOR]

The South China Sea is China's largest marginal sea area, and it is rich in oil and gas mineral resources; thus, estimating its sea level changes is of practical significance. Based on linear and nonlinear sea level change characteristics, this paper decomposes 1992–2019 monthly mean sea level anomaly time series in the South China Sea into trend, seasonal, and random terms. This paper compares the seasonal autoregressive integrated moving average (SARIMA) and Prophet models for estimating the trend and seasonal terms and the long short-term memory (LSTM) and radial basis function (RBF) models for estimating random terms, and the more suitable models were selected. A Prophet-LSTM combined model was developed based on the accuracy results. This paper uses the combined model to study the effect of known data length on the experimental results and determines the best prediction duration. The results show that the combined model is suitable for short-term and medium-term estimations of 12–36 months. The accuracy at 36 months is 0.962 cm, which proves that the combined model has high application value for estimating sea level changes in the South China Sea. [ABSTRACT FROM AUTHOR]

Modeling interpretable artificial intelligence (AI) for flood forecasting represents a serious challenge: both accuracy and interpretability are indispensable. Because of the uncertainty and nonlinearity of flood, existing hydrological solutions always achieve low prediction robustness while machine learning (ML) approaches neglect the physical interpretability of models. In this paper, we focus on the need for flood forecasting and propose an interpretable Spatio-Temporal Attention Long Short Term Memory model (STA-LSTM) based on LSTM and attention mechanism. We use dynamic attention mechanism and LSTM to build model, Max-Min method to normalize data, variable control method to select hyperparameters, and Adam algorithm to train the model. Emphasis is placed on the visualization and interpretation of attention weights. Experiment results on three small and medium basins in China suggest that the proposed STA-LSTM model outperforms Historical Average (HA), Fully Connected Network (FCN), Convolutional Neural Networks (CNN), Graph Convolutional Networks (GCN), original LSTM (LSTM), spatial attention LSTM (SA-LSTM), and temporal attention LSTM (TA-LSTM) in most cases. Visualization and interpretation of spatial and temporal attention weights reflect the reasonability of the proposed attention-based model. [ABSTRACT FROM AUTHOR]

A major difficulty in integrating large scale wind power generation in an electrical power system is that wind generated power appears to be erratic, intermittent, and volatile. In this paper, we demonstrate the efficacies of a novel ultra short term 1-step ahead wind generated power prediction model, by combining two best of breed machine learning models in their respective areas of applications: a deep convolutional neural network (CNN) model, known to be effective in classification problems, and a bi-directional long short term memory (Bi-LSTM) model, known to be effective in 1-step ahead time series prediction problems, using a cross attention (XA) mechanism on three challenging practical datasets: the East-China dataset, the Yalova (Turkey) dataset, and the 16 MW dataset. There are two alternative cross attention models: (1) using the CNN features as the key–value pair, and the Bi-LSTM features as the query, this we called a XA model, and (2) using the Bi-LSTM features as the key–value pair, and the CNN features as the query, this we called a XA_I model; we found empirically for a properly chosen sliding window length, whichever CNN model or Bi-LSTM model performs well in the 1-step ahead prediction, then its features should be used as the query in the cross attention mechanism, thus, the XA achieves better results than XA_I on the East-China dataset, while XA_I achieves better results than XA on the Yalova dataset, and on the 16 MW dataset. Where comparable results using other state-of-the-art models are available on the Yalova dataset and the 16 MW dataset, we found that our XA_I results are superior. Moreover, it is found empirically, that the cross attention mechanism model produces stable k -step ahead predictions, for k ≤ W 2 , where W is the sliding window length. • A novel prediction model for wind generated power based on cross attention mechanism has been proposed. • The proposed method demonstrated on three practical datasets to have superior performance. • The cross attention mechanism based models give a stable multi-step ahead prediction. [ABSTRACT FROM AUTHOR]

Due to the complexity and volatility of stock market trading, there are still some issues in the existing prediction methods, including the processing of data noise, inexplicable selection of model parameters, and the lag in predicting price fluctuations. Aiming to better predict daily stock prices and fluctuations caused by high noise non-stationary data in actual trading, we propose a hybrid deep learning framework based on Singular Spectrum Analysis (SSA), multiple feature selection, and Long Short Term Memory (LSTM) network optimized by Particle Swarm Optimization (PSO). Based on Pearson correlation coefficients, we select features highly correlated with the closing price as inputs, and further achieve the noise reduction and optimization of those input sources by applying SSA to decompose the stock price time series into independent component signals. Using the stock price data from China and USA, we compared the prediction performance of our method with several well-known methods, and found that it achieved higher price prediction accuracy during periods of high stock volatility. For example, the R-squared and prediction accuracy of Shanghai Composite Index achieved 0.9998 and 99.01%, while the prediction metrics of S &P 500 reached 0.9883 and 94.26%, respectively. Besides, considering transaction costs, our method also achieved the highest profit in trading tests, even compared to long-term holding strategy. [ABSTRACT FROM AUTHOR]

The non-US countries' oil trade will be affected by the combined fluctuations in oil price and exchange rate. For China, the world's largest oil importer, any slight fluctuations in oil price or exchange rate will significantly impact his oil import. Therefore, this paper firstly quantified the oil price spread between nominal price and real price corresponding to its domestic currency to show such impact. Secondly, we conducted statistical tests on the price spread under the circumstance when China imports from Russia, Oman, Dubai, WTI and Brent, and we introduced the fractal model to explore the nonlinear characteristics of the five price spread series. The five series all showed obvious long-term memory characteristics. Thirdly, time-varying memory series were constructed to further explore their dynamic features and the results showed they could forewarn the reversal trend of price spread series. Finally, the complex network is applied to analyze the dynamical mechanism of time-varying memory series and further explore the laws of the price spread. The algorithm could provide references for policymakers to anticipate the general trend of price spread and compare the price advantage of the crude oil. • The combined volatility of crude oil price and exchange rate affects trade cost. • We explored the long-term memory of the crude oil import payment pressure index. • The sliding window was applied to explore the time-varying memory. • We detected the evolution characteristics and mechanism of the time-varying memory. [ABSTRACT FROM AUTHOR]

With the opening of the third energy revolution, the booming development of the natural gas industry in China has led to the existing pipeline network facilities no longer meeting the gas demand of its residents. How to adjust pipeline network layout and design peak-regulating gas storage facilities to meet the gas demand of residents has become a challenge. However, accurate load forecasting can help solve this problem. To that end, this paper proposes a novel combined model that integrates local mean decomposition (LMD), wavelet threshold denoising (WTD), and long short term memory (LSTM) methods to predict short-term gas consumption, or "LMD-WTD-LSTM". To begin with, based on the above three basic methods, this paper adopts two different combination schemes, schemeⅠand schemeⅡ, to predict the gas consumption of London city. The results show that schemeⅡ is better than schemeⅠ. Further, this paper compares the predicted results with four other forecasting methods: such as back propagation neural network (BPNN) model, the gated recurrent unit (GRU) model, the multiple linear regression (MLR) model, and the LSTM model. Finally, this paper discusses the influence of different temperatures and different forecasting time lengths (FTL) on the prediction accuracy of the LMD-WTD-LSTM model. The results show that when the forecasting time length is 20 days, the mean absolute percentage error (MAPE) of schemeⅡ is 11.63 %, which is the lowest in comparison to the other four methods. According to this study, the LMD-WTD-LSTM method has an excellent performance in daily natural gas prediction. • A method, local mean decomposition-wavelet threshold denoising-long short term memory is proposed. • Wavelet threshold denoising can remove the outliers caused by accidents. • Local mean decomposition simplifies the complexity of the original load. • Proper variables in a certain region can help improve accuracy. [ABSTRACT FROM AUTHOR]

One of the biggest difficulties, which China's power infrastructure is now experiencing, is the capacity to estimate demand of short-term electricity. It can accurately estimate changes in the total power load in particular locations. Accurate forecasting results, which also serve as a dependable guide for power system operation, may increase the flexibility and resource usage of the contemporary power market. Power load characteristics are impacted by several things. This study presents an estimate method for short-term demand of electricity based on WOA-BILSTM in order to comprehensively evaluate the time series features presented in the power load data and improve the accuracy of power load forecasting. After extracting features for the variables that influence power load, this method uses the bidirectional long-term and short-term memory (BILSTM) neural network layer for bidirectional time series feature learning. Using the local electricity load data from Quanzhou from 2018 as the data set, the prediction model is constructed by screening the multi-dimensional input parameters and selecting the feature vectors with good association as the input carefully. By comparing the outcomes of three popular load forecasting models, including LSTM network, BILSTM network, and WOA-BILSTM, it is demonstrated that the WOA-BILSTM neural network method is more accurate and effective than others two. By building an optimal combination model, this technique could increase the accuracy of power load data's short-term predictions. At the same time, it also reduces the time of personnel debugging. [ABSTRACT FROM AUTHOR]

Microelectromechanical Systems (MEMS) Inertial Measurement Unit (IMU) containing a three-orthogonal gyroscope and three-orthogonal accelerometer has been widely utilized in position and navigation, due to gradually improved accuracy and its small size and low cost. However, the errors of a MEMS IMU based standalone Inertial Navigation System (INS) will diverge over time dramatically, since there are various and nonlinear errors contained in the MEMS IMU measurements. Therefore, MEMS INS is usually integrated with a Global Positioning System (GPS) for providing reliable navigation solutions. The GPS receiver is able to generate stable and precise position and time information in open sky environment. However, under signal challenging conditions, for instance dense forests, city canyons, or mountain valleys, if the GPS signal is weak and even is blocked, the GPS receiver will fail to output reliable positioning information, and the integration system will fade to an INS standalone system. A number of effects have been devoted to improving the accuracy of INS, and de-nosing or modelling the random errors contained in the MEMS IMU have been demonstrated to be an effective way of improving MEMS INS performance. In this paper, an Artificial Intelligence (AI) method was proposed to de-noise the MEMS IMU output signals, specifically, a popular variant of Recurrent Neural Network (RNN) Long Short Term Memory (LSTM) RNN was employed to filter the MEMS gyroscope outputs, in which the signals were treated as time series. A MEMS IMU (MSI3200, manufactured by MT Microsystems Company, Shijiazhuang, China) was employed to test the proposed method, a 2 min raw gyroscope data with 400 Hz sampling rate was collected and employed in this testing. The results show that the standard deviation (STD) of the gyroscope data decreased by 60.3%, 37%, and 44.6% respectively compared with raw signals, and on the other way, the three-axis attitude errors decreased by 15.8%, 18.3% and 51.3% individually. Further, compared with an Auto Regressive and Moving Average (ARMA) model with fixed parameters, the STD of the three-axis gyroscope outputs decreased by 42.4%, 21.4% and 21.4%, and the attitude errors decreased by 47.6%, 42.3% and 52.0%. The results indicated that the de-noising scheme was effective for improving MEMS INS accuracy, and the proposed LSTM-RNN method was more preferable in this application. [ABSTRACT FROM AUTHOR]

• A grey seasonal trend support vector machine for energy forecasting is proposed. • The model can efficiently identify the seasonal fluctuations in energy data. • The model is powerful to predict China's energy production. • The predictive performance of the new model is much superior to the benchmarks. This paper proposed a hybrid ensemble forecasting technique that incorporates the merits of the accumulation generation operation (AGO), least-square support vector regression (LSSVR), dummy variable, and time trend item to forecast the seasonal time series characterized by nonlinearity and uncertainty. This proposed framework, a grey seasonal trend least squares support vector machine (GSTLSSVR), can make it easy to add arbitrary seasonality to any feature, which improves model realism significantly. Furthermore, the grid search method and the last block validation are developed to identify the optimal hyperparameter, making it possible for the developed method to capture seasonal and nonlinear patterns of the original data. For illustration and verification, the quarterly petroleum coke production and monthly crude oil production in China are employed to construct diverse competing models, including the seasonal fluctuation grey model (SGM(1,1)), seasonal autoregressive integrated moving average (SARIMA), Holt-Winters (HW), backpropagation neural network (BPNN), a grey seasonal least squares support vector machine (GSLSSVR), long short term memory (LSTM) and LSSVR. The forecasting values of quarterly petroleum coke production are 668.74, 682.68, 679.99, 704.35 for 2019 and 717.59, 734.99, 727.16, 741.53 for 2020, with the lowest MAPEP value (3.7566%). Moreover, the monthly predicted crude oil production for 2019 are 3113.9, 1645, 1562.2, 1625.4, 1599.3, 1619.1, 1627, 1576.9, 1629.6, 1583.6, and 1615.6, while those counterparts for 2020 are 3093.2, 1625, 1547.5, 1621.5, 1610.6, 1644.9, 1662.9, 1615.4, 1667.4, 1615, 1667.4, 1615, and 1640.2, with the lowest MAPEP value (1.0034%). Generally, the empirical results show that the proposed model obtains superior overall performance to other benchmarks in both case studies. Therefore, this proposed model exhibits promising prospects for identifying seasonal patterns in energy production sequences. [ABSTRACT FROM AUTHOR]

As a representative of the process industry, the cement industry consumes a large amount of coal and electricity resources. This is mainly due to the irrational energy scheduling caused by the rough production and independent statistics of each energy consumption indicator within the cement industry. The synchronous accurate prediction of energy consumption can provide a more resultful scheme for the production control process and energy scheduling. However, due to the time delay, variables coupling, and uncertainty of production, it is difficult to synchronously forecast multiple indicators. In this paper, a data-driven prediction method combining Sliding Window and Dual-Channel Convolutional Neural Network (SWDC-CNN) is proposed to achieve synchronous prediction of coal and electricity consumption in the next 5 min. The sliding window method is designed to extract time-varying delay characteristics of time series data to overcome its influence on energy consumption prediction. The effect of redundant parameters between weakly correlated variables on energy prediction is reduced by designing a dual-channel structure. We experimented and compared with excellent models Support Vector Machine (SVM), eXtreme Gradient Boosting (XGBoost), Recurrent Neural Network (RNN), Long Short Term Memory (LSTM), and Gate Recurrent Unit (GRU) on an actual cement production data in Shanxi Province, China. Experimental results show that the proposed SWDC-CNN model has good performance, the highest prediction accuracy, and can meet the expected requirements. • A synchronous energy consumption prediction model consisting of sliding window and dual-channel CNN is proposed. • Design a sliding window to construct a high-dimensional time series matrix containing time-varying delay features. • Dual-channel convolutional neural network is designed to solve the problem of parameter redundancy. • The model can accurately predict coal and electricity consumption of cement clinker calcination process simultaneously. [ABSTRACT FROM AUTHOR]

The largest earthquake (Ms 8½) in eastern China occurred near Tancheng in 1668. Six M ≥ 6 earthquakes occurred in this region in historical time. We re-examine the primary historical archives from this region, adding independent constraints such as maps of active faults, local topography, intensity attenuation laws for China, and relevant administrative practices in archiving data, to reassess source parameters (epicenter, magnitude, or maximum intensity). We use the seismic density index—a measure of the clustering of seismicity weighted by source magnitude—to identify candidate zones of present-day clustering of seismicity than may represent a long-term memory of past historical events in an intraplate setting. We update at least one of the source parameters for four of the seven events. The main shock and two aftershocks of the 1668–1672 Ms 8½ Tancheng earthquake sequence locate along the strike of the Tan-Lu fault zone, each associated with significant and contiguous seismic density anomalies whose elliptical contours are elongated along its strike. A third major aftershock (Ms ≈ 6) in this sequence locates some 200 km away from the fault zone. Six of the seven historical events have seismic density anomalies that cannot be explained by aftershocks of modern seismicity (ML ≤ 5.3), and hence are candidate zones for long-lived aftershock sequences. In contrast, there are no significant seismic density anomalies along the section of Tan-Lu fault south of Suqian, despite palaeo-seismic evidence for a Ms 8½ earthquake around 6280 B.C., providing an upper bound to long-term memory of great earthquakes in this region. [ABSTRACT FROM AUTHOR]

In recent years, air pollution has become a factor that cannot be ignored, affecting human lives and health. The distribution of high-density populations and high-intensity development and construction have accentuated the problem of air pollution in China. To accelerate air pollution control and effectively improve environmental air quality, the target of our research was cities with serious air pollution problems to establish a model for air pollution prediction. We used the daily monitoring data of air pollution from January 2016 to December 2020 for the respective cities. We used the long short term memory networks (LSTM) algorithm model to solve the problem of gradient explosion in recurrent neural networks, then used the particle swarm optimization algorithm to determine the parameters of the CNN-LSTM model, and finally introduced the complete ensemble empirical mode decomposition of adaptive noise (CEEMDAN) decomposition to decompose air pollution and improve the accuracy of model prediction. The experimental results show that compared with a single LSTM model, the CEEMDAN-CNN-LSTM model has higher accuracy and lower prediction errors. The CEEMDAN-CNN-LSTM model enables a more precise prediction of air pollution, and may thus be useful for sustainable management and the control of air pollution. [ABSTRACT FROM AUTHOR]

The proliferation of spam in China has a negative impact on internet users' experiences online. Existing methods for detecting spam are primarily based on machine learning. However, it has been discovered that these methods are susceptible to adversarial textual spam that has frequently been imperceptibly modified by spammers. Spammers continually modify their strategies to circumvent spam detection systems. Text with Chinese homophonic substitution may be easily understood by users according to its context. Currently, spammers widely use homophonic substitution to break down spam identification systems on the internet. To address these issues, we propose a Bidirectional Gated Recurrent Unit (BiGRU)–Text Convolutional Neural Network (TextCNN) hybrid model with joint embedding for detecting Chinese spam. Our model effectively uses phonetic information and combines the advantages of parameter sharing from TextCNN with long-term memory from BiGRU. The experimental results on real-world datasets show that our model resists homophone noise to some extent and outperforms mainstream deep learning models. We also demonstrate the generality of joint textual and phonetic embedding, which is applicable to other deep learning networks in Chinese spam detection tasks. [ABSTRACT FROM AUTHOR]

The monitoring and prediction of the groundwater level (GWL) significantly influence the landslide kinematics. Based on the long-term fluctuation characteristics of the GWL and the time lag of triggering factors, a dynamic prediction model of the GWL based on the Maximum information coefficient (MIC) algorithm and the long-term short-term memory (LSTM) model was proposed. The Sifangbei landslide in the Three Gorges Reservoir area (TGRA) in China, wherein eight GWL monitoring sensors were installed in different locations, was taken as a case study. The monitoring data represented that the fluctuation of the GWL has a specific time lag concerning the accumulated rainfall (AR) and the reservoir water level (RWL). In addition, there were spatial differences in the fluctuation of the GWL, which was controlled by the elevation and the micro landform. From January 19, 2015, to March 6, 2017, the measured data were used to set up the predicted models. The MIC algorithm was adopted to calculate the lag time of the GWL, the RWL, and the AR. The LSTM model is a time series prediction algorithm that can transmit historical information. The Gray wolf optimization (GWO) algorithm was used to seek the most suitable hyperparameter of the LSTM model under the specific prediction conditions. The single-factor GWO-LSTM model without considering triggering factors and the support vector machine regression (SVR) model were considered to compare the prediction results. The results indicate that the MIC-GWO-LSTM model reached the highest accuracy and improved the prediction accuracy by considering the factor selection process with the learner training process. The proposed MIC-GWO-LSTM model combines the advantages of each algorithm and effectively constructs the response relationship between the GWL fluctuation and triggering factors; it also provides a new exploration for the GWL prediction, monitoring, and early warning system in the TGRA. [ABSTRACT FROM AUTHOR]

• A Convolutional Long Short Term Memory Network is used to predict the flood events based on deep learning techniques. • The spatial and time characteristics of floods in China are well modeled to overcome the shortcomings generated by merely relying on time-series analysis. • Different from traditional methods, the hydrological area is gridded into different watersheds for future processing using image processing methods. Floods cause substantial damage across the world every year. Accurate and timely prediction of floods can significantly minimize the loss of life and property. Recently, numerous machine learning models have been used for flood prediction, showing that their performance is preferable to traditional statistical models. However, the existing models neglect the spatial features of floods, which drive flood generation and concentration. In this paper, the area of interest is divided into grids based on longitude and latitude, and the rainfall and discharge collected by stations are combined into tensors according to station coordinates. Different from one-dimensional time series, our input feature is a two-dimensional time series with spatial information. Hence, combining a Convolutional Neural Network (CNN) with a Long Short Term Memory Network (LSTM), we propose the convolution LSTM (ConvLSTM) to extract spatiotemporal features of hydrological information. The methodology is demonstrated using the hydrological data collected at the Xi County stations, located on the Huai River in Henan Province, China. Numerical results indicate that the relative error of arrival time is within 30%, and the relative error of peak discharge is within 20%, satisfying the 2005 Chinese Water Resource Standard on flood prediction permit error. The experiments also show that the ConvLSTM outperforms the recent models in terms of flood arrival time and peak discharge, thereby proving a promising alternative. [ABSTRACT FROM AUTHOR]

Accurate wave forecasts provide vital safety guarantees for marine operations of ships. The difficulty in the study of ocean wave forecasting is that ocean waves have non-linear and non-stationary characteristics. The Auto-Regressive (AR) model is widely used in time series forecasting, but it is weak in processing non-linearity and non-stationarity. The Long Short Term Memory (LSTM) model developed for timing problems has strong non-linear processing capabilities, but there are shortcomings in the treatment of non-stationarity. Empirical Mode Decomposition (EMD) can effectively separate non-linearity and non-stationarity in data. This paper combines the advantages of the LSTM model and EMD and proposes an EMD-LSTM method. Based on significant wave heights from three locations offshore China, it has been found that the error of the EMD-LSTM method is lower than that of the LSTM model. Meanwhile, it has been found that the EMD-LSTM method can increase the forecast time in advance by more than doubled under the same tolerance by analyzing the forecast effect of different forecast times. It has been proved that the EMD-LSTM model has good superiority for the prediction of non-linear and non-stationary waves. • Combining EMD algorithm and LSTM model to forecast non-stationary waves. • The improvement of the EMD-LSTM method compared with the LSTM model is analyzed. • The influence of wave non-stationary characteristics and advance forecast time on forecast performance. • Verification and comparison studies are carried out using actual sea measurement data sets. [ABSTRACT FROM AUTHOR]

Based on the data ranging from the first quarter of 2000 to the fourth quarter of 2009, this paper, by adopting the FIGARCH model, empirically studies whether there exists a long-term memory fluctuation rate of housing price. The findings indicate that the fluctuation of China's housing price is characterized by long-term memory, agglomeration effect and heteroscedasticity. Some suggestions are then put forward on the regulation and control of real estate and housing price--more attention should be paid to quantitative tools in housing price regulation, the policy stability and durability should be emphasized, and the financial institutions should evaluate the credit risks of real estate irregularly. [ABSTRACT FROM AUTHOR]

Reliable and accurate streamflow forecasting plays a vital role in the optimal management of water resources. To improve the stability and accuracy of streamflow forecasting, a hybrid decomposition-ensemble model named VMD-LSTM-GBRT, which is sensitive to sampling, noise and long historical changes of streamflow, was established. The variational mode decomposition (VMD) algorithm was first applied to extract features, which were then learned by several long short-term memory (LSTM) networks. Simultaneously, an ensemble tree, a gradient boosting tree for regression (GBRT), was trained to model the relationships between the extracted features and the original streamflow. The outputs of these LSTMs were finally reconstructed by the GBRT model to obtain the forecasting streamflow results. A historical daily streamflow series (from 1/1/1997 to 31/12/2014) for Yangxian station, Han River, China, was investigated by the proposed model. VMD-LSTM-GBRT was compared with respect to three aspects: (1) feature extraction algorithm; ensemble empirical mode decomposition (EEMD) was used. (2) Feature learning techniques; deep neural networks (DNNs) and support vector machines for regression (SVRs) were exploited. (3) Ensemble strategy; the summation strategy was used. The results indicate that the VMD-LSTM-GBRT model overwhelms all other peer models in terms of the root mean square error (RMSE = 36.3692), determination coefficient (R2 = 0.9890), mean absolute error (MAE = 9.5246) and peak percentage threshold statistics (PPTS(5) = 0.0391%). The addressed approach based on the memory of long historical changes with deep feature representations had good stability and high prediction precision. [ABSTRACT FROM AUTHOR]

Oil has to be redistributed around the world because of their uneven distribution. Therefore, the method of accurately identifying and forecasting the risks of oil import has always been the focus of research. Thus, we re-examined the risk of oil import from the whole process of oil import. Based on resilience theory, a framework for risk assessment was established by referring to the 4 A factor (availability, accessibility, affordability and acceptability). Then, long and short term memory network model (LSTM) was constructed and trained to forecast the risk of oil import. Taking the oil import network in China as an example, by comparing with the BP, SVM and CNN model, the better fitting effect and higher forecasting accuracy of LSTM model were verified; According to the results, from 2011 to 2018, China's oil import system was less resilient and experienced different stages, which are driven by different dominate factors. Besides, availability and affordability risks remain severe in the foreseeable future. Therefore, China should optimize the combination of exporters, actively participate in the development of transportation routes, establish and improve China's crude oil futures market, and plan the layout in advance to avoid oil import risks. • This paper examined the risk of oil import from the whole process. • A standardized framework for risk assessment was established. • The LSTM model was constructed to forecast the risk of oil import. [ABSTRACT FROM AUTHOR]

This paper applies the extended MFDFA method to analyze the volatility characteristics, multifractal features and asymmetry in EU carbon emission permits trading market(EUA) and the main Chinese carbon emission permits market including HBEA, SZA and GDEA. We divided the EUA market into three stages according to the phased construction plan of European commission. The empirical analysis shows that the fluctuation in the second and third stage of EUA market presents a long-term memory feature. However, there is obvious anti-persistent characteristics in the volatility of three main Chinese carbon emission permits trading markets and the first stage in EUA market. The multifractal degree in three main Chinese carbon emission permits trading markets is less than that in the first stage of EUA market, but far greater than the second and third stages. We also verified that the HBEA market has the weakest volatility and the least multifractal degree among three Chinese markets, and whose validity is second to the third stage of EUA market. In addition, the fluctuation of the EU and Chinese carbon emission permits trading markets are asymmetric. The EUA market has stronger reaction to good news, while Chinese carbon emission permits trading market still displays the leverage effect which verifies the volatility is more sensitive to the bad news in the second and third stages. Therefore, the results powerfully demonstrate that the Chinese carbon emission permits trading market is less mature and highly dynamic, which is similar to the level of first stage of EUA market. The volatility in Chinese markets is strong, and exists much noise in the system. So, the efficiency in Chinese markets is weaker than EU. Chinese markets are still far from maturity and efficiency. • The market price volatility does not possess the significant long-term memory characteristics, it shows up the anti-persistent features. • With the development of the EUA market, the multifractal characteristics become weaker and weaker, and the reaction of the market price fluctuation to information become weaker as well. • The market asymmetry reflects the reaction of the price volatility to the positive and negative information. [ABSTRACT FROM AUTHOR]

It is well recognized that climate predictability has three origins: (i) climate memory (inertia of the climate system) that accumulated from the historical conditions, (ii) responses to external forcings, and (iii) dynamical interactions of multiple processes in the climate system. However, how to systematically identify these predictable sources is still an open question. Here, we combine a recently developed Fractional Integral Statistical Model (FISM) with a Variance Decomposition Method (VDM), to systematically estimate the potential sources of predictability. With FISM, one can extract the memory component from the considered variable. For the residual parts, VDM can then be applied to extract the slow varying covariance matrix, which contains signals related to external forcings and dynamical interactions of multiple processes in climate. To demonstrate the feasibility of this new method, we analyzed the seasonal predictability in observational monthly surface air temperatures over China from 1960 to 2017. It is found that the climate memory component contributes a large portion of the seasonal predictability in the temperature records. After removing the memory component, the residual predictability stems mainly from teleconnections, i.e., in summer the residual predictability is closely related to sea surface temperature anomalies (SSTA) in the eastern tropical Pacific and the northern Indian Ocean. Our results offer the potential of more skillful seasonal predictions compared with the results obtained using FISM or VDM alone. [ABSTRACT FROM AUTHOR]

This paper aims to use recurrent neural networks (RNNs) to perform landslide susceptibility mapping in Yongxin County, China. The two main contributions of this study are summarized as follows. First, the regular RNN is compared to its three variants in the case study of landslide susceptibility mapping for the first time, including long short term memory, gated recurrent unit and simple recurrent unit. Second, a sequential data representation method is proposed to fully explore the predicting potential of RNNs. The study area consists of 364 historical landslide locations that were divided into two parts: 255 (70%) for training and 109 (30%) for validation, and 16 landslide influencing factors were considered for spatial prediction. To validate the effectiveness of these RNN-related methods, several objective measures of accuracy, recall, F-measure, Matthews correlation coefficient and the receiver operating characteristic were used for evaluation. Experimental results demonstrate that very high and high susceptible areas are concentrated in the northwest and south of Yongxin County, while landslides in the central area are less prone to occur. Based on quantitative results, all the RNN-related methods achieved area under the curve values above 0.83 and produced accurate prediction results with the optimized parameters. Therefore, the RNN framework can be used as a useful tool for the landslide susceptibility mapping task to mitigate and manage landslides. • Landslide susceptibility mapping based on recurrent neural networks. • Sequential landslide data are constructed for modelling. • Regular recurrent neural network is compared with its three variants. • Reliable landslide susceptibility maps can be achieved in the study area. [ABSTRACT FROM AUTHOR]

Long-term climate memory is ubiquitous in climate systems, but its contribution to climate prediction has not been assessed systematically. We used an integral fractional statistical model (FISM) to quantify climate memories in different variables over China. Their contributions to climate prediction were estimated using explained variances. We found different climate memory effects for different variables in different regions. The contribution of climate memory to climate variability is stronger in temperature than in precipitation records. For temperatures (including both air temperature and land temperature), the average variance explained by climate memory is around 3∼4%. For precipitation, the average explained variance was 0.6%. The low values for explained variances indicate that, on average, the contributions of climate memory to temperature and precipitation predictions are small. But in specific regions, higher climate memory effects may occur. For precipitation, climate memory can contribute 3% of the variance in southeast China. For temperature, climate memory can explain ≥ 10% of the variance in northeast and southwest China, which is not low and should be considered in prediction. Therefore, for more accurate climate prediction, we suggest first determining the contribution of climate memory. For variables or regions with strong climate memory effects, a scheme considering climate memory effects may help improve future climate predictions. [ABSTRACT FROM AUTHOR]

It has been well recognized that, for most climatic records, their current states are influenced by both past conditions and current dynamical excitations. However, how to properly use this idea to improve the climate predictive skills, is still an open question. In this study, we evaluated the decadal hindcast experiments of 11 models (participating in phase 5 of the Coupled Model Intercomparison Project, CMIP5) in simulating the effects of past conditions (memory part, M(t)) and the current dynamical excitations (non-memory part, ε (t) ). Poor skills in simulating the memory part of surface air temperatures (SAT) are found in all the considered models. Over most regions of China, the CMIP5 models significantly overestimated the long-term memory (LTM) of SAT. While in the southwest, the LTM was significantly underestimated. After removing the biased memory part from the simulations using fractional integral statistical model (FISM), the remaining non-memory part, however, was found reasonably simulated in the multi-model means. On annual scale, there were high correlations between the simulated and the observed ε (t) over most regions of the country, and for most cases they had the same sign. These findings indicated that the current errors of dynamical models may be partly due to the unrealistic simulations of the impacts from the past. To improve predictive skills, a new strategy was thus suggested. As FISM is capable of extracting M(t) quantitatively, by combining FISM with dynamical models (which may produce reasonable estimations of ε (t) ), improved climate predictions with the effects of past conditions properly considered may become possible. [ABSTRACT FROM AUTHOR]

Honeybees are valuable pollinators strongly influencing ecosystem conservation, stability, genetic variation and ecological relationships in the floral diversity, plant community, evolution, and specialization. Honeybee is an excellent model organism for research on learning and memory among invertebrates. In their behavior, the subjective evaluation of a sucrose stimulus influences the behavioral performance. Here we report comparative behavioral data on the sensitization, habituation, short-term and long-term memory of Apis cerana and A. mellifera in China, using different sucrose concentrations. A. mellifera foragers have higher sucrose responsiveness than A. cerana foragers when tested using a proboscis extension response (PER) assay. Sensitization and habituation are well-known forms of nonassociative learning. A. cerana took less number of habituation trials as compared to A. mellifera. Thus, these results significantly showed that A. cerana took less time to dishabituate. The sensitive stimuli of A. cerana against lemon extract were more than A. mellifera. In addition, although A. mellifera showed more sensitive stimuli against apple extract than A. cerana cerana, A. mellifera displayed significantly more learning and memory behavior than A. cerana after 2 and 24 hours. [ABSTRACT FROM AUTHOR]

In this work, multifractal methods have been successfully used to characterize the temporal fluctuations of daily Jiuzhai Valley domestic and foreign tourists before and after Wenchuan earthquake in China. We used multifractal detrending moving average method (MF-DMA). It showed that Jiuzhai Valley tourism markets are characterized by long-term memory and multifractal nature in. Moreover, the major sources of multifractality are studied. Based on the concept of sliding window, the time evolutions of the multifractal behavior of domestic and foreign tourists were analyzed and the influence of Wenchuan earthquake on Jiuzhai Valley tourism system dynamics were evaluated quantitatively. The study indicates that the inherent dynamical mechanism of Jiuzhai Valley tourism system has not been fundamentally changed from long views, although Jiuzhai Valley tourism system was seriously affected by the Wenchuan earthquake. Jiuzhai Valley tourism system has the ability to restore to its previous state in the short term. [ABSTRACT FROM AUTHOR]